| Along with the development of economy,a lot of highway projects in our country have been developed,more and more projects had to face the construction under complex geological conditions.The highway tunnel was an important part of the highway system.More and more tunnels were built in gypsum rock strata.Gypsum rock was a kind of typical expansive corrosion surrounding rock,its corrosion would cause serious deterioration of the lining structure strength,and affect the safety of the tunnel.In view of this engineering disease,this paper took Lirang tunnel as the engineering support,the effects of different factors on the strength of concrete after corrosion were studied by orthogonal test.Based on the BP neural network,the prediction model of concrete corrosion strength change was established.The numerical simulation of the stress of tunnel lining concrete after corrosion was carried out.The main research contents and conclusions were as follows:(1)Orthogonal test was used to study the corrosion characteristics of concrete under the combined action of concrete impermeability grade,C3 A content,corrosion solution pressure,SO42-concentration and time.The effects of different factors on the strength of concrete after corrosion were analyzed.the results showed: Quadratic functional relation existed between anti permeability level,C3 A content and concrete corrosion strength,which was a logarithmic function of sulfate concentration or time,however pressure and concrete corrosion strength are approximate linear relationship;(2)The influence weight of different factors on the strength of concrete after corrosion was analyzed qualitatively and quantitatively using the range analysis method and the analytic hierarchy process.The results of range analysis showed that the weight of each factor was in accordance with the following: C3 A content,sulfate concentration,time,impermeability grade,solution pressure;the results of the analytic hierarchy process showed that the weight of each factor was 0.3797,0.319,0.101,and 0.0186,respectively.(3)Based on BP neural network,The prediction model of the change of the strength of the concrete after the corrosion of the concrete with the number of hidden layer neurons was 3 ~ 13,and then it was trained and tested.The prediction results of the model were the most accurate when the number of neurons in the hidden layer was 9,the maximum error of the five predicted samples was 4.2%.(4)Based on the principle of damage mechanics,thickness of concrete corrosion completely model was established.Based on the predicted value of the predicted model of concrete strength after corrosion,it was calculated that the thickness of the lining structure was 7.36 cm,8.10 cm,8.54 cm,8.85 cm,9.10 cm,9.30 cm,9.48 cm,9.63 cm,9.76 cm,9.88 cm when the lirang tunnel operating period of 10 years,20 years,......100 years later.(5)Based on the conventional load-structure model,a load-structure model based on contact analysis was proposed using the contact analysis method.Using this method,Lirang tunnel was simulated,the maximum compressive stress and maximum tensile stress variation in the Lirang tunnel operation period was obtained.(6)From several aspects measures to improve the tunnel lining structure by sulfate erosion was put forward.It was proposed to improve the corrosion resistance of shotcrete,to improve the corrosion resistance of the two lining,and to adopt the whole waterproof board,water repellent material and so on.Under the optimized parameters,the prediction model of the strength change of concrete after corrosion was used to predict the strength change.The improvement of the strength change was 35%... |
PDF Full Text Request